Identifikasi Kerusakan Permukaan Jalan Menggunakan Artificial Intelligence (Ai) Sebagai Dasar Klasifikasi Jenis Kerusakan Berdasarkan Pedoman Bina Marga (Studi Kasus: Jalan Kelapapati Laut,Kecamatan Bengkalis)

Road surface damage is one of the problems that can reduce road comfort, safety, and serviceability, requiring an accurate identification process as a basis for determining
the type of damage. The use of Artificial Intelligence is an alternative approach to assist in the automatic identification of road damage through survey images. This study aims to identify road surface damage using Artificial Intelligence, analyze the identification results, and use the identification results as a basis for classifying damage types based on the Bina Marga Guidelines. The research was conducted on Jalan Kelapapati Laut, Bengkalis Regency, through field surveys, road damage image acquisition, dataset development, and YOLOv8 model training using Roboflow and Google Colab. The results showed that out of 186 research images, 141 images were successfully identified, while 45 images were not successfully identified. From the successfully identified images, 673 damage objects were obtained, consisting of 412 Alligator Crack objects, 197 Pothole objects, and 64 Patching objects. Based on field validation results, the damage severity consisted of 81 data categorized as Light Damage (R), 58 data as Moderate Damage (S), and 47 data as Severe Damage (B). The results indicate that Artificial Intelligence can be used as a supporting tool for identifying road surface
damage; however, the identification results still require validation against field conditions and observation parameters based on the Bina Marga Guidelines.

Keywords: Artificial Intelligence, road damage identification, YOLOv8, road surface damage, Bina Marga Guidelines